Application of Kernel-Based Feature Space Transformations and Learning Methods to Phoneme Classification

This paper examines the applicability of some learning techniques to the classification of phonemes. The methods tested were artificial neural nets (ANN), support vector machines (SVM) and Gaussian mixture modeling (GMM). We compare these methods with a traditional hidden Markov phoneme model (HMM), working with the linear prediction-based cepstral coefficient features (LPCC). We also tried to combine the learners with linear/nonlinear and unsupervised/supervised feature space transformation methods such as principal component analysis (PCA), independent component analysis (ICA), linear discriminant analysis (LDA), springy discriminant analysis (SDA) and their nonlinear kernel-based counterparts. We found that the discriminative learners can attain the efficiency of HMM, and that after the transformations they can retain the same performance in spite of the severe dimension reduction. The kernel-based transformations brought only marginal improvements compared to their linear counterparts.     

VPM: A visual,precise and multilevel metamodeling framework for describing mathematical domains and UML (The Mathematics of Metamodeling is Metamodeling Mathematics)

As UML 2.0 is evolving into a family of languages with individually specified semantics, there is an increasing need for automated and provenly correct model transformations that (i) assure the integration of local views (different diagrams) of the system into a consistent global view, and, (ii) provide a well-founded mapping from UML models to different semantic domains (Petri nets, Kripke automaton, process algebras, etc.) for formal analysis purposes as foreseen, for instance, in submissions for the OMG RFP for Schedulability, Performance and Time. However, such transformations into different semantic domains typically require the deep understanding of the underlying mathematics, which hinders the use of formal specification techniques in industrial applications. In the paper, we propose a multilevel metamodeling technique with precise static and dynamic semantics (based on a refinement calculus and graph transformation) where the structure and operational semantics of mathematical models can be defined in a UML notation without cumbersome mathematical formulae.     

Noise peaks influence communication in the operating room. An observational study

Noise peaks are powerful distractors. This study focuses on the impact of noise peaks on surgical teams’ communication during 109 long abdominal surgeries. We related measured noise peaks during 5-min intervals to the amount of observed communication during the same interval. Results show that noise peaks are associated with less case-relevant communication; this effect is moderated by the level of surgical experience; case-relevant communications decrease under high noise peak conditions among junior, but not among senior surgeons. However, case-irrelevant communication did not decrease under high noise level conditions, rather there was a trend to more case-irrelevant communication under high noise peaks. The results support the hypothesis that noise peaks impair communication because they draw on attentional resources rather than impairing understanding of communication. As case-relevant communication is important for surgical performance, exposure to high noise peaks in the OR should be minimised especially for less experienced surgeons.

Practitioner Summary: This study investigated whether noise during surgeries influenced the communication within surgical teams. During abdominal surgeries, noise levels were measured and communication was observed. Results showed that high noise peaks reduced the frequency of patient-related communication, but did not reduce patient-irrelevant communication. Noise may negatively affect team coordination in surgeries.     

Is my model any good: differentially private regression diagnostics

Linear and logistic regression are popular statistical techniques for analyzing multi-variate data. Typically, analysts do not simply posit a particular form of the regression model, estimate its parameters, and use the results for inference or prediction. Instead, they first use a variety of diagnostic techniques to assess how well the model fits the relationships in the data and how well it can be expected to predict outcomes for out-of-sample records, revising the model as necessary to improve fit and predictive power. In this article, we develop \(\epsilon \)-differentially private diagnostics tools for regression, beginning to fill a gap in privacy-preserving data analysis. Specifically, we create differentially private versions of residual plots for linear regression and of receiver operating characteristic (ROC) curves as well as binned residual plot for logistic regression. The residual plot and binned residual plot help determine whether or not the data satisfy the assumptions underlying the regression model, and the ROC curve is used to assess the predictive power of the logistic regression model. These diagnostics improve the usefulness of algorithms for computing differentially private regression output, which alone does not allow analysts to assess the quality of the posited model. Our empirical studies show that these algorithms can be effective tools for allowing users to evaluate the quality of their models.     

Impact of growth rate on the quality of ZNS-MQW InGaAsP/InP laser structures grown by LP-MOVPE

We investigated the influence of the growth rate on the quality of zero-net-strained InGaAsP/InGaAsP/InP multiquantum well structures for 1.55 μm emission grown by low pressure metalorganic vapor phase epitaxy. The samples consisted of fixed compressive strained wells (ɛ=+1%) and tensile strained barriers (ɛ=−0.5%) grown with different quaternary bandgap wavelengths (λ_B=1.1–1.4 μm). Using higher growth rates, we obtained for the first time high quality zero net strained multi quantum well structures, regardless having constant group V composition in the well and barriers. The samples were analyzed by x-ray diffraction, photoluminescence and atomic force microscopy techniques. The amplitude of surface modulation roughness along [011] direction decreased from 20 nm to 0.53 nm with increasing growth rate and/or quaternary compositions grown outside the miscibility gap. A new deep PL broad emission band strongly correlated with the onset of wavy layer growth is also reported. Broad area and ridge waveguide lasers with 10 wells exhibited low losses (34 cm⁻¹) and low threshold current densities at infinite cavity length (1020 A·cm⁻² and 1190 A·cm⁻², respectively).     

An ensemble-based system for microaneurysm detection and diabetic retinopathy grading

Reliable microaneurysm detection in digital fundus images is still an open issue in medical image processing. We propose an ensemble-based framework to improve microaneurysm detection. Unlike the well-known approach of considering the output of multiple classifiers, we propose a combination of internal components of microaneurysm detectors, namely preprocessing methods and candidate extractors. We have evaluated our approach for microaneurysm detection in an online competition, where this algorithm is currently ranked as first, and also on two other databases. Since microaneurysm detection is decisive in diabetic retinopathy (DR) grading, we also tested the proposed method for this task on the publicly available Messidor database, where a promising AUC 0.90 ± 0.01 is achieved in a "DR/non-DR"-type classification based on the presence or absence of the microaneurysms.     

Variational optical flow computation in real time.

This paper investigates the usefulness of bidirectional multigrid methods for variational optical flow computations. Although these numerical schemes are among the fastest methods for solving equation systems, they are rarely applied in the field of computer vision. We demonstrate how to employ those numerical methods for the treatment of variational optical flow formulations and show that the efficiency of this approach even allows for real-time performance on standard PCs. As a representative for variational optic flow methods, we consider the recently introduced combined local-global method. It can be considered as a noise-robust generalization of the Horn and Schunck technique. We present a decoupled, as well as a coupled, version of the classical Gauss-Seidel solver, and we develop several multgrid implementations based on a discretization coarse grid approximation. In contrast, with standard bidirectional multigrid algorithms, we take advantage of intergrid transfer operators that allow for nondyadic grid hierarchies. As a consequence, no restrictions concerning the image size or the number of traversed levels have to be imposed. In the experimental section, we juxtapose the developed multigrid schemes and demonstrate their superior performance when compared to unidirectional multgrid methods and nonhierachical solvers. For the well-known 316 x 252 Yosemite sequence, we succeeded in computing the complete set of dense flow fields in three quarters of a second on a 3.06-GHz Pentium4 PC. This corresponds to a frame rate of 18 flow fields per second which outperforms the widely-used Gauss-Seidel method by almost three orders of magnitude.     

An end‐to‐end alternative to TCP PEPs: Initial Spreading,a TCP fast start‐up mechanism

While Transmission Control Protocol (TCP) Performance Enhancing Proxy (PEP) solutions have long been undisputed to solve the inherent satellite problems, the improvement of the regular end‐to‐end TCP congestion avoidance algorithms and the recent emphasis on the PEPs drawbacks have opened the question of the PEPs sustainability. Nevertheless, with a vast majority of Internet connections shorter than ten segments, TCP PEPs continue to be required to counter the poor efficiency of the end‐to‐end TCP start‐up mechanisms. To reduce the PEPs dependency, designing a new fast start‐up TCP mechanism is therefore a major concern. But, while enlarging the Initial Window (IW) up to ten segments is, without any doubt, the fastest solution to deal with a short‐lived connection in an uncongested network, numerous researchers are concerned about the impact of the large initial burst on an already congested network. Based on traffic observations and real experiments, Initial Spreading has been designed to remove those concerns whatever the load and type of networks. It offers performance similar to a large IW in uncongested network and outperforms existing end‐to‐end solutions in congested networks. In this paper, we show that Initial Spreading, taking care of the satellite specificities, is an efficient end‐to‐end alternative to the TCP PEPs. Copyright © 2015 John Wiley & Sons, Ltd.     

Free-scaling your data center

The increasing popularity of both small and large private clouds and expanding public clouds poses new requirements to data center (DC) architectures. First, DC architectures should be incrementally scalable allowing the creation of DCs of arbitrary size with consistent performance characteristics. Second, initial DC deployments should be incrementally expandable supporting small-scale upgrades without decreasing operation efficiency. A DC architecture possessing both properties satisfies the requirement of free-scaling.Recent work in DC design focuses on traditional performance and scalability characteristics, therefore resulting in symmetric topologies whose upgradability is coarse-grained at best. In our earlier work we proposed Scafida, an asymmetric, scale-free network inspired DC topology which scales incrementally and has favorable structural characteristics. In this paper, we build on Scafida and propose a full-fledged DC architecture achieving free-scaling called FScafida. Our main contribution is threefold. First, we propose an organic expansion algorithm for FScafida; this combined with Scafida’s flexible original design results in a freely scalable architecture. Second, we introduce the Effective Source Routing mechanism that provides near-shortest paths, multi-path and multicast capability, and low signaling overhead by exploiting the benefits of the FScafida topology. Third, we show based on extensive simulations and a prototype implementation that FScafida is capable of handling the traffic patterns characteristic of both enterprise and cloud data centers, tolerates network equipment failures to a high degree, and allows for high bisection bandwidth.